Complex traits are inherited via the combined effects of quantitative trait loci segregating in a population. These genetic determinants may have large or small effects, and may combine in a myriad of ways. Despite over a decade of work on mapping the genetic determinants of complex traits via genome-wide association studies and other methods, the field has as yet not determined an optimal approach to inferring phenotype from genotype. We propose to use yeast genetics to learn the underlying genetic architecture of a series of quantitative traits controlling stress tolerance. The yeast system has superior experimental tools, conserved biology across eukaryotes, and a growing collection of diverse genome sequences on which to draw. We will first focus on loci of strong effect, i.e. ?Mendelian?traits. These lend themselves to easy genetic mapping and experimental confirmation via allele swap experiments. Our hypothesis is that genes with strong effect alleles will also harbor alleles of more quantitative effect across a population. This idea is akin to the ?rare variant? hypothesis in which low frequency strong effect alleles contribute to high disease risk. We seek to discover whether these same genes may also be of importance across the population due to lower effect size alleles that may be more difficult to identify. We will pursue this project in three specific aims as part of an integrated team with expertise in yeast genetics, genomics, and genome evolution. In Aim 1, we will establish a diallel pairwise offspring panelto investigate the genetic underpinnings of traits. For this purpose, we will perform pairwise crosses, generate full meiotic offspring for each cross, and phenotype the parental isolates, hybrids, and the offspring using high throughput methods. Then, we will estimate the genetic complexity of traits by analyzing their inheritance patterns. This approach will also clearly identify traits showing simple, Mendelian inheritance. The second aim is to determine the genetic basis of those phenotypes that segregate as a small number of large effect loci. To map the causative loci, we will use a bulk segregant mapping method coupled with deep sequencing, followed by reciprocal hemizygosity and/or allele swap experiments to prove causation. The third aim is to test whether these same loci are important for phenotypic variation across a large panel of strains. We will amplify alleles from >1000 genetically diverse isolates, transplant them into multiple genetic backgrounds, and assay gene function via a pool-based, quantitative, competition assay. Allele frequency will be measured using deep sequencing. Once we have identified alleles with differential functional consequences, we will use a novel application of DNA shuffling to map the causative polymorphisms to single base resolution. This combination of methods will allow us to identify crosses in which stress tolerance traits segregate in a genetically simple manner, map the causative genes and nucleotides, and determine the importance of additional variants in these genes across a population. Our results will inform the understanding of the genetic basis of complex traits at a scale never before possible.